Project Summary Use of contraception for pregnancy prevention is among the most important behaviors affecting fertility in a population, and determining the trajectory of individual lives. But, contraceptive failures contribute to unintended, and often unwanted, conceptions. Almost all U.S. women use contraception at some point in their lives and millions of individuals rely on scientifically-valid estimates of the risk of failure. Clinicians also rely on these key measures to counsel clients in their choice of method as well as to establish broad practice guidelines with wide-reaching influence. And, because of the high quality of U.S. population data, contraceptive failure rates based on the experience of U.S. women are used in international research and policy applications. Indeed, few other products of demographic research are as widely used as contraceptive failure rates. In spite of their significance to public health and research, no new methodological advances have been offered in this area of work in almost 30 years and virtually no research exists to illuminate factors that account for variation in the risk of contraceptive failure across individuals, populations, locations or time. To address limitations of prior studies, improve the quality of failure rate estimates and forge a new path for research, we propose five linked aims. Aim 1 will advance understanding of factors that contribute to differential experiences with contraceptive failure by developing a novel theoretical framework specifically focused on individuals' perceptions, beliefs, relationships and external influences on the risk of contraceptive failure. We will exploit a wealth of existing qualitative data to inform development of the framework and then design and analyze new survey questions to test that framework in a large nationally representative survey of reproductive age women. In Aim 2, we will advance the measurement of contraceptive failure by developing a new approach for estimating failure rates using Bayesian statistics to more fully utilize existing knowledge and to surpass prior approaches that have limited our understanding of variation in failure rates across population groups. In Aim 3, we will use these new subgroup estimates to identify the factors that led to change over time in method- specific failure rates and the role of more effective use of contraception in recent declines in U.S. unintended pregnancy rates. This work has important implications not only for changes in the U.S. population, but could also inform new approaches to examining the impact of contraceptive use worldwide. In Aim 4, we expand our investigation to structural influences on contraceptive failure by examining the extent to which state policy climates and the socioeconomic characteristics of their residents are associated with variation in contraceptive failure for women who share similar demographic traits but live in different states. In Aim 5, we will work with an advisory panel of providers and leaders in reproductive health care to develop effective dissemination of a new set of contraceptive failure rates for clinical and public use.